Start Submission Become a Reviewer

Reading: Common sources of occupant dissatisfaction with workspace environments in 600 office buildings

Download

A- A+
Alt. Display

Research

Common sources of occupant dissatisfaction with workspace environments in 600 office buildings

Authors:

Thomas Parkinson,

Center for the Built Environment (CBE), University of California—Berkeley, Berkeley, CA, US
X close

Stefano Schiavon ,

Center for the Built Environment (CBE), University of California—Berkeley, Berkeley, CA, US
X close

Jungsoo Kim,

School of Architecture, Design and Planning, The University of Sydney, Sydney, NSW, AU
X close

Giovanni Betti

Center for the Built Environment (CBE), University of California—Berkeley, Berkeley, CA, US
X close

Abstract

Previously unpublished data from over 600 office buildings in the Center for the Built Environment (CBE) Occupant Survey database are used to perform a systematic analysis of dissatisfaction in contemporary workspaces. A total of 81% of respondents expressed dissatisfaction with at least one aspect of their workspace, and 67% with more than one. Acoustics were the most common source of dissatisfaction, particularly related to people talking, speech privacy, and phones. Other challenges included a perceived lack of control over the temperature and insufficient space, along with other associated problems of densely populated offices. The analysis shows that context matters when understanding occupant dissatisfaction. Occupants of open-plan offices with low or no partitions were almost twice as likely to complain about their workspace than someone in a private, enclosed office. Being near a window decreased the likelihood of dissatisfaction compared with those who were not near a window. There was a clear relationship between self-perceived performance and satisfaction with the indoor environment. Dissatisfaction profiles found that acoustics, space, and privacy-related items co-occur for many occupants dissatisfied with more than one workspace aspect.

 

Practical relevance

Post-occupancy surveys are a useful tool for evaluating whether an office environment supports occupants while conducting their work. While highlighting the successes is important, complaints from dissatisfied occupants can identify issues and pinpoint reasons why spaces do not meet expectations. The reported challenges generally relate to the simultaneous reduction in control and personalization with increasingly open and densely populated layouts. Occupant dissatisfaction may impact performance given the reported relationship between satisfaction with the environment and feeling supported by the workspace to complete work tasks. The themes emerging from this analysis identify common dissatisfaction sources that can serve as an empirical basis to identify common problems in contemporary workspace designs.

How to Cite: Parkinson, T., Schiavon, S., Kim, J., & Betti, G. (2023). Common sources of occupant dissatisfaction with workspace environments in 600 office buildings. Buildings and Cities, 4(1), 17–35. DOI: http://doi.org/10.5334/bc.274
523
Views
128
Downloads
8
Twitter
  Published on 04 Jan 2023
 Accepted on 19 Dec 2022            Submitted on 09 Oct 2022

1. Introduction

Post-occupancy evaluations (POEs) are a systematic process of evaluating the occupant experience of a building (Preiser et al. 2015). They are a common tool in measuring the success of workspace designs (Li et al. 2018) and informing the operation and management of buildings by providing feedback for diagnostic purposes. A feedback loop established through POE can improve the fit between buildings and their users (Zimmerman & Martin 2001) by focusing on occupant needs in addition to objective metrics of building performance (e.g. energy, water, waste, transportation). They are also widely used by research disciplines interested in workspace satisfaction (Li et al. 2018). Of the different POE methods for soliciting occupant feedback (Dykes & Baird 2013), the questionnaire survey is most widely used because it is simple and cost-effective (Heinzerling et al. 2013).

POE questionnaires are recommended as a ‘basic’ instrument for subjective data collection in building performance evaluation protocols from professional and regulatory bodies such as the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE), US Green Building Council (USGBC), and The Chartered Institution of Building Services Engineers (CIBSE). Recent growth in survey deployments has been driven largely by international building certification schemes such as Leadership in Energy and Environmental Design (LEED), WELL Building Standard, Building Research Establishment Environmental Assessment Method (BREEAM), and Green Star. Standardized POE databases enable comparisons between buildings, organizations, or design features that often form the basis of performance metrics and benchmarks. These resources have been extensively analysed by researchers to generate a snapshot of occupant satisfaction in buildings. Some examples include green and non-green buildings (Altomonte et al. 2017; Altomonte & Schiavon 2013), different heating, ventilation, and air-conditioning (HVAC) system types (Brager & Baker 2009; Karmann et al. 2017; Kim & de Dear 2012), different spatial configurations (Kim & de Dear 2013; Leder et al. 2016), determinants of overall satisfaction (Frontczak et al. 2012), and occupant demographics (Choi et al. 2010; Kim et al. 2013).

POE studies to date have revealed important patterns and phenomena across the commercial building stock, but the real-world impact is often muted by limited knowledge transfer from academia to industry. Generalized information from metanalyses can be difficult for practitioners to operationalize and apply to building management practices. Notable exceptions include periodic reports from commercial entities such as the Leesman Review and Gensler US Workplace Survey. These are valued by practitioners but receive less traction within the research discipline of POEs. Individual case studies can also be overlooked in academic papers because the specific context makes it difficult to generalize any findings (Moezzi & Goins 2011). Furthermore, POE studies have generally focused on occupant satisfaction without necessarily using additional POE questions that further contextualize the satisfaction votes (Dutta et al. 2020; Kim et al. 2016; Zagreus et al. 2004).

Bridging the gap between the POE research discipline and actionable insights for practitioners to improve workspaces (e.g. facilities managers and building operators) may require a shift in focus from the determinants of overall satisfaction to common sources of occupant dissatisfaction. The authors have argued (Graham et al. 2021) that POE surveys should examine the positive attributes of workspace designs as well as the poor-performing pieces. An earlier analysis of the Center for the Built Environment (CBE) Occupant Survey database found that dissatisfied occupants are valuable sources of information when evaluating both the successes and the failures of workspace design attributes (Kent et al. 2021). Dissatisfied occupants can identify issues that may have gone unnoticed and can assist building managers in pinpointing the reasons why spaces do not meet expectations. But this data source often lacks necessary details for practitioners to properly diagnose problems (McArthur et al. 2018). Without a systematic understanding of common sources of dissatisfaction, building operators rely on ad-hoc response to individual complaints that often result in temporary remedies rather than focused resolutions (Goins & Moezzi 2013).

Identifying common sources of occupant dissatisfaction is a necessary foundation in helping guide proactive building management practices to improve the workspace’s performance. In this paper, data from over 600 office buildings in the CBE Occupant Survey database are used to better understand occupant dissatisfaction. This dataset is previously unpublished and complements the extant literature stemming from the CBE Occupant Survey (e.g. Graham et al. 2021). The aim of the present study is to quantify common sources of dissatisfaction in office buildings, explore the contextual factors that may influence the source of dissatisfaction, and investigate the interrelationships between common sources of dissatisfaction.

2. Methods

This analysis of occupant dissatisfaction with office workspaces is based on responses to the CBE Occupant Survey, a long-running POE survey used in over 800 buildings. Graham et al. (2021) provide a recent overview of the tool and summary statistics for the database. The survey asks respondents to evaluate their satisfaction with 15 workspace items and two overall items (building and personal workspace). Figure 1 shows a summary of respondent satisfaction with those items. Branching questions are only shown to dissatisfied respondents and include a pre-populated list of checkbox items designed to identify the reasons for dissatisfaction (see Figure A1 in Appendix A for an example of the branching question format used in the CBE Occupant Survey). These branching questions are included for eight of the 15 items: acoustics (sound privacy or noise level), air quality, office layout (amount of space or ease of interaction or and visual privacy), lighting, and temperature. This category of data is referred to here as ‘dissatisfaction sources.’ Data-cleaning included removing surveys in buildings other than offices, those with few responses (fewer than 20), or instances where a modified survey for a specific research project was used. The final dataset had responses from 62,360 occupants in 617 office buildings in the US (83%), Australia (6%), Canada (5%), and other locations (6%). All analyses were done in R (v. 4.2) and RStudio (v. 1.4.1103) with the ‘tidyverse’ (Wickham et al. 2019) suite of packages.

Summary of satisfaction votes (seven-point Likert scale) for the 17 items in the Center for the Built Environment (CBE) Occupant Survey
Figure 1 

Summary of satisfaction votes (seven-point Likert scale) for the 17 items in the Center for the Built Environment (CBE) Occupant Survey.

Note: Items are arranged from lowest satisfaction (top) to highest satisfaction (bottom). The percentages of all satisfied votes (green) and all dissatisfied votes (red) are reported for each item.

Source: Modified after Graham et al. (2021).

The modular structure of the CBE Occupant Survey meant that some restructuring of the data was needed for the different analyses. To explore how contextual factors influence the likelihood of dissatisfaction using logistic regression, the dissatisfaction sources were combined with the six available background questions (gender, age, years in the building, time in the role, type of workspace, and proximity to a window). In addition, seven-point Likert questions about self-reported performance (‘does ___ enhance or interfere with your ability to get your job done?’) were paired with responses to the relevant branching question to determine the relationship between those two questions. The dissatisfaction sources were converted into a logical format (0 = no dissatisfaction source/satisfied, 1 = dissatisfaction source/dissatisfied) and the self-reported performance responses were collapsed into three categories (< 0 = interferes, 0 = neither, > 0 = enhances) to simplify the results by showing a general relationship between these factors. These data were used to calculate the proportions of responses across the related self-reported performance data on a 100-repeated balanced random sample (n = 4000 per performance vote, 12,000 total responses). This additional step was used to ensure that the results were robust by removing any sampling bias. The overall aim of comparing indoor environmental quality (IEQ) and performance was to highlight the anticipated relationship between dissatisfaction with the environment and respondents’ self-reported productivity.

While the CBE Occupant Survey is a distinct dataset, mining POE databases is relatively common in building science. As such, sharing details for reproducibility is relevant to interpreting the results as well as encouraging similar analysis for comparison with the findings reported here. The ‘base R’ implementation of binomial logistic regression was used to model the probability of dissatisfaction across any of the seven categories. The exponentiated odds ratios were calculated across significant background variables to determine the likelihood of dissatisfaction. This transformation is commonly used to make interpretation easier. Extending this to explore the dissatisfaction sources, ‘dissatisfaction profiles’ were defined using cluster analysis based on the tallied number of sources for the seven categories. Using the ‘cluster’ package implementation of k-means (v. 2.1.1; Maechler et al. 2021), every respondent was assigned membership to one of five clusters. Silhouette values (Rousseeuw 1987) were used to determine the two- and five-cluster solutions as optimum and focused on the five-cluster solution for greater granularity. The cluster analysis was used to find some overarching groups within the diverse set of respondents that are somewhat generalizable to other contexts and environments.

Determining the clusters is a useful method to group the respondents, but in itself does not reveal the reason for the clustering. Recursive partitioning of background variables using the rpart package (v. 4.1-15; Therneau & Atkinson 2019) was used to help explain cluster membership. The overly dissatisfied clusters were dropped on the assumption they were indiscriminately dissatisfied with their space. They also represented a very small portion (3%) of the overall dataset, so doing so helped subsequent under-sampling to balance across the four remaining clusters. Hyperparameters for the decision tree were a minimum split of 100, a minimum bucket of 30, a maximum depth of 5, and a complexity parameter of 0.001. An 80/20 split was used to estimate model accuracy.

The final analysis using association rule-learning complements the earlier findings by exploring relationships between multiple dissatisfaction sources within the database rather than just individual results. The rationale is that understanding occurrences of multiple dissatisfaction sources that appear frequently may reveal common and systemic challenges in contemporary work environments. The a priori algorithm in the arules package (v. 1.6-6; Hahsler et al. 2021) was applied to a logical matrix of dissatisfaction sources to find the 10 most frequent rules (support of 0.005) for each category. The minimum rule length was 3, the maximum rule length was 6, and only rules with a confidence > 0.75 were included.

3. Results

There was a total of 124,602 identified dissatisfaction sources in the CBE Occupant Survey database. Approximately 81% of survey respondents were dissatisfied with at least one aspect of their workspace, and 67% with two or more. The proportion of respondents who chose at least one of the checkbox items for the seven categories is shown in Figure 2. Most sources were about acoustics (54% of respondents), followed by temperature (38%) and visual privacy (28%). Ease of interaction had the lowest rate (10%). The percentage of respondents who provided a reason for their dissatisfaction was slightly below the total percentage dissatisfied. This mirrors dissatisfaction votes reported in Figure 1 and is expected given the branching structure of the satisfaction questions in the survey.

Percentage of respondents who gave at least one reason for their dissatisfaction
Figure 2 

Percentage of respondents who gave at least one reason for their dissatisfaction.

Note: The number of dissatisfied respondents for each category is shown alongside the percentages.

3.1 Dissatisfaction sources

Overall dissatisfaction rates are a useful diagnostic and indicate problem areas for office buildings. However, branching questions in the CBE Occupant Survey provide greater insights because respondents may specify the reasons for dissatisfaction. The list of reasons in the survey is predefined and the available options vary between categories. Text fields are also available and are collapsed into ‘Other.’ Responses to the branching questions are shown in Figure 3. Most dissatisfaction sources for workspace acoustics related to people talking, speech privacy, and phones. This suggests that the main distraction is from potentially intelligible speech by occupants rather than equipment noise or outdoor sounds. For thermal comfort, a perceived lack of control over the temperature and spatial variations in temperature were the main problems. Air movement being too high was more frequent than too low, but similar percentages reflect the diverse thermal preferences of an office worker population (Parkinson et al. 2021). Dissatisfaction with lighting was evenly split between those wanting less and those wanting more light in their workspace. Smells from food was the most common dissatisfaction source related to air quality, but the large number of ‘other’ responses suggests there are issues beyond those included in the checklist.

Dissatisfaction sources for the seven categories with branching questions in the survey
Figure 3 

Dissatisfaction sources for the seven categories with branching questions in the survey.

Note: Percentages show the share of dissatisfied respondents within each category. Respondents could select more than one source, so they do not total to 100%. Colors indicate different categories.

Dissatisfaction sources in related categories of office layout—amount of space, ease of interaction, visual privacy—revolve around insufficient space and the associated problems with densely populated offices. For example, the largest number of dissatisfied respondents within the amount of space category said there was not enough room at their individual work areas (74%). The main privacy-related concerns within the visual privacy category were passers-by (55%) and partition height being too low (40%). Similarly, collegial interaction is commonly hampered by concerns of noise distracting others (46%), as well as limited casual spaces (41%) and opportunities for interaction (30%). These issues relate to the design and configuration of offices rather than the IEQ per se.

3.2 Dissatisfaction and performance

An occupant may be dissatisfied with an aspect of their workspace but still feel that it supports them in their work tasks. The proportion of respondents nominating at least one dissatisfaction source was calculated to test the relationship between dissatisfaction with the physical environment and self-reported performance. Figure 4 shows that respondents identifying a dissatisfaction source are much more likely to feel that the workspace is interfering with their ability to get their job done. Conversely, satisfied respondents comprise the large majority who say their workspace enhances their ability to get their job done. The difference in proportions between the questions suggests that some aspects of the workspace impinge upon self-reported performance more than others. For example, the relationship between dissatisfaction and self-reported performance was strongest for acoustics, then temperature and lighting. Respondents were less likely to connect air quality dissatisfaction sources to self-reported performance.

Proportion of respondents selecting at least one dissatisfaction source across the self-reported performance metric
Figure 4 

Proportion of respondents selecting at least one dissatisfaction source across the self-reported performance metric.

Note: The dataset was balanced (under-sampled) across the responses to the self-reported productivity question and the proportions calculated using random samples repeated 100 times. Only items related to the physical environment with both the performance and branching dissatisfaction questions were included.

3.3 Contextual factors of dissatisfaction

The likelihood of dissatisfaction with workspace environments was determined using exponentiated odds ratios from binomial logistic regressions with responses to the background questions as independent variables. Six background variables were tested, with the type of workspace, years in building, and proximity to a window being significant (p < 0.001). Demographic factors of age, gender, and time in the role were not significant when the three significant factors were considered. This may be due to missing responses (age = 41% missing, gender = 10% missing) given the sensitivity of the questions, collinearity with other questions (e.g. time in the role), or a lack of explanatory power. The binomial logistic regression reported in Figure 5 demonstrates that the odds of dissatisfaction are influenced by both personal and workspace factors. The type of workspace (enclosed or open) had the largest influence on the likelihood of dissatisfaction with any of the seven categories. Occupants of open-plan offices with low or no partitions were almost twice as likely to complain about their workspace than someone in a private, enclosed office. There was a general increase in dissatisfaction with the number of years spent working in the building; someone working for more than five years was 28% more likely to be dissatisfied than someone there for less than one year. Finally, being near a window (within 5 m) decreased the likelihood of dissatisfaction compared with those who were not near a window.

Exponentiated odds ratios of a respondent being dissatisfied with any of the seven categories with branching questions based on background variables
Figure 5 

Exponentiated odds ratios of a respondent being dissatisfied with any of the seven categories with branching questions based on background variables.

Note: Grey bars report the 95% confidence interval and colors indicate the different background variables. Reference levels are shown with smaller bars to help with the comparison of odds ratios for the other responses.

3.4 Dissatisfaction profiles

The influence of background factors on the likelihood of dissatisfaction begged the question if subpopulations had distinct patterns or profiles of dissatisfaction sources. For example, are those respondents dissatisfied with the lack of privacy also dissatisfied with acoustics and cleanliness? These insights may better help diagnose issues and assist in determining the appropriate architectural, technological, and/or policy solutions to simultaneously address multiple problems. To determine these ‘dissatisfaction profiles,’ cluster analysis was used to define the five groups shown in Figure 6 based on the dissatisfaction sources. There is an inverse trend between cluster size and the median number of dissatisfaction sources. The largest is cluster 1, which groups respondents with the fewest dissatisfaction sources. Cluster 2 differs from cluster 1 due to the higher number of acoustic issues (median of four acoustic dissatisfaction sources). Along with acoustic concerns, members of cluster 3 were also dissatisfied with their visual privacy. Cluster 3 is labelled as those having issues with workspace privacy. Cluster 4 is similar but also includes dissatisfaction with the amount of space. Lastly, cluster 5 is respondents who are generally dissatisfied and selected multiple sources across all categories.

‘Dissatisfaction profiles’ of survey respondents determined using cluster analysis
Figure 6 

‘Dissatisfaction profiles’ of survey respondents determined using cluster analysis.

Note: Clusters are ordered from the fewest (left) to the most complaints (right) and were given descriptive labels summarizing the nature of dissatisfaction. Colors indicate different categories, and brightness is based on the median number of dissatisfaction sources (shown as text labels; the text is suppressed if the median was zero). The number of respondents in each cluster and the unique survey count are reported at the bottom.

Cluster analysis is an unsupervised machine-learning technique to assign group memberships within a dataset based on distance or (dis)similarity. It is useful in exploratory data analysis but here it does not explain the type of occupants belonging to each cluster. Recursive partitioning was used to build the decision tree shown in Figure 7 to provide insights for classifying cluster membership based on background variables. This analytical technique is used to find groups within datasets for classification. Cluster 5 was removed as those respondents may have been expressing general dissatisfaction with their workplace rather than about specific IEQ or layout issues. Low model accuracy (33%) suggests the background variables in the survey do not sufficiently capture the diversity of factors contributing toward the reason for occupant dissatisfaction. However, it does offer some general insights into the dissatisfaction profiles established through the cluster analysis. The main factor distinguishing cluster membership is the type of office (enclosed or open). These split respondents into those with few dissatisfaction sources (cluster 1) in enclosed offices or high partitions, and those with privacy issues (cluster 3) in open offices with low or no partitions. Occupants of private offices were most likely to have few dissatisfaction sources (cluster 1). Those with high partitions experienced acoustic issues (cluster 2), while those in shared offices were dissatisfied with privacy and space (cluster 4). In most cases, occupants in open offices with the low or no partitions had privacy issues (cluster 3). The only exception was those without partitions who were not near a window and had been working in the building for more than five years (cluster 4). Gender and age were considered in the recursive partitioning but did not improve model classification beyond that of the other parameters.

Decision tree classifying cluster membership using recursive partitioning on the balanced (under-sampled) dataset
Figure 7 

Decision tree classifying cluster membership using recursive partitioning on the balanced (under-sampled) dataset.

Note: The decision tree starts at the root node (top) and with each classification will branch into subsequent decision nodes (rectangles) before terminating at the leaf nodes (bottom). Branches are the background variables and nodes are the most likely cluster after each split. Final cluster memberships are shown in the leaf nodes along with their labels.

3.5 Associations between dissatisfaction sources

Finally, the relationships between dissatisfaction sources and co-occurrences across the different categories assessed by the occupant survey were explored. Association rule-mining of the dissatisfaction sources found 140,658 rules meeting the support and confidence criteria (see the methods section for criteria). Item frequencies are summarized in Figure A2 in Appendix A and are dominated by acoustic issues such as people talking and phones. The 10 most frequent rules for each category are shown as a parallel coordinates plot in Figure 8. The most frequent rule—{speech privacy, people talking on the phone => people talking}—occurred 14,911 times. Acoustic items are present in all the most frequent rules except for the lighting category. Frequent rules also contain at least one other closely related item from the same category. For example, ‘area hotter/colder than others’ and ‘no control of thermostat’ in the temperature category, or ‘too much light’ and ‘reflections’ in the lighting category.

Parallel coordinates plot of the 10 most frequent rule associations for each category
Figure 8 

Parallel coordinates plot of the 10 most frequent rule associations for each category.

Note: Association rule-mining is used to find patterns in large databases with many different parameters. Each line traces common connections between dissatisfaction sources in different categories. For example, the pink line shows that insufficient space in work areas and meeting rooms (Amount of Space) commonly occurs with people talking (Acoustics) despite them being from different categories. The heavy colored lines are the most frequent rule for each category (color), and the light grey lines are the nine remaining frequent rules. The specific dissatisfaction sources are shown on the right and the categories to which they belong are shown on the left. Following each line will show common patterns between dissatisfaction sources; the direction of the line is not important.

Association rule-mining reveals common connections across categories when occupants are dissatisfied with more than one item. For example, the issue of insufficient work area (amount of space) frequently occurs with a lack of casual spaces (ease of interaction), people talking on the phone (acoustics), and concern about noise being distracting for others. These dissatisfaction sources are from different questions but frequently occur together when respondents are dissatisfied with visual privacy, ease of interaction, amount of space, and acoustics. Similarly, the problem of passers-by (visual privacy) co-occurs with insufficient space in work areas (amount of space) and people talking (acoustics). These dissatisfaction sources are distinct items but are thematically linked as issues of space and privacy. Association rule-mining reveals connections between items in different categories that might be overlooked in large and complex datasets such as the CBE Occupant Survey database.

4. Discussion

The principal purpose of an office is to provide an environment that supports occupants in achieving their work tasks to meet the broader goals of the employer. Delivering on this promise includes several key professional industries. The physical attributes of an office fall under the purview of workspace designers and interior architects. IEQ, on the other hand, is more the responsibility of architects and engineers. All these aspects, however, are only partially under the control of the organization that occupies the space. For example, a tenant may control desk layout, partitions, ergonomics, and occupant behavior (to an extent) but may not have any influence over window views, daylighting, or the overall thermal environment if they are renting. And building operation is often the responsibility of the building owners and/or their contracted facilities management companies. A successful workspace design therefore requires close collaboration between different stakeholders. Such synchronicity is difficult to achieve in real-world settings in the commercial property sector. For this reason, identifying common sources of dissatisfaction across a diverse collection of offices provides important background knowledge to inform new office designs and future retrofit projects.

The analysis shows a clear relationship between self-perceived performance and satisfaction with the indoor environment (Figure 4). This evidence should motivate organizations to respond proactively to occupant dissatisfaction and prioritize the improvement of indoor environments. The results of the logistic regression (Figure 5) demonstrate that context matters when understanding occupant dissatisfaction. Many of the factors contributing to the likelihood of complaint are related but distinct from the physical attributes of the space. For example, the time someone has worked in a building has a significant effect on the likelihood of dissatisfaction. While it is difficult to parse the reasons for this in the dataset, it suggests other potentially confounding factors such as job satisfaction (Cheung et al. 2022) or workplace culture (De Been & Beijer 2014) influence workspace satisfaction over time. Alternatively, there may be a ‘newness’ effect that leads occupants to overlook issues initially but wears off over time to reveal problem areas. This phenomenon is often referred to as the ‘honeymoon effect’ (Fichman & Levinthal 1991).

The finding that factors beyond the indoor environment have some bearing on the likelihood of dissatisfaction demonstrates the importance of understanding the occupant and their needs in combination with other possible physical contributors to workspace satisfaction. The industry around commercial office space tends to focus much more on the physical attributes of the space—furniture, layout, IEQ—than the human resources occupying them. The results of the analysis, specifically the logistic regression (Figure 5), cluster analysis (Figure 6), and decision tree (Figure 7), suggest a strong influence coming from the occupant as well as the space. This underscores the need for POEs to better understand the people occupying spaces, e.g. by assessing their life and job satisfaction. It is also a reminder for practitioners in the architecture, engineering, and construction industry and their clients to make the environment fit the occupant rather than forcing the occupant to fit the environment. These ideas are expanded on in a companion paper stemming from the same dataset (Graham et al. 2021).

4.1 Interventions

The size of the CBE Occupant Survey database makes this a systematic analysis of dissatisfaction sources in contemporary workspaces. It is impossible to reduce this multidimensional dataset sourced from multiple contexts into directly actionable items applicable in every workspace. Yet there is appeal in further developing the dissatisfaction profiles (Figure 6), decision tree (Figure 7), and rule associations (Figure 8) as a diagnostic tool. For example, a building manager receiving complaints about something from the occupants could possibly use these analyses to identify and respond to the potential source of the problem. Furthermore, such workflows could be implemented into computer-aided facility management platforms or integrated workplace management systems to improve the operation and maintenance of buildings (Burak Gunay et al. 2019). Future work may consider exploring this potential further.

Themes throughout the analyses identify common dissatisfaction sources that are relevant to many commercial office buildings. When combined with contextual factors such as occupant traits and workspace configuration, these high-level findings provide an empirical basis from which to identify common problems in contemporary workspace designs. Most of these challenges have been reported in the extant literature from field studies or POEs in office buildings. That body of research is too large and diverse to summarize here completely. Instead, the following sections outline common interventions for practitioners aiming to reduce occupant complaints based on select research findings and professional judgment. Focusing on the environment is a pragmatic approach, but future work should expand on this with greater emphasis on the occupant.

4.1.1 Office layout

There has been discussion for several decades about the advantages and disadvantages of open-plan offices (e.g. Oldham & Brass 1979; Brennan et al. 2002; Bernstein & Turban 2018). The analysis shows clear evidence that open-plan office layouts lead to more occupant complaints than private offices (Figure 5). More recently, activity-based workplaces have emerged as a solution to balancing the challenges and benefits of open offices. Activity-based workplaces designs have seen continued uptake and are purported to be more satisfactory across several measures compared with open plan (Engelen et al. 2019). While it remains an active research topic, it is generally accepted that understanding how workspace environments best support occupants requires nuanced evaluations and a clear articulation of the measured outcomes (Vischer 2008).

It is not the present authors’ belief that open offices are intrinsically bad and do not serve occupants in any context. Rather, there are a greater number of considerations that designers need to make to implement open office layouts successfully when compared with private offices. For example, the regression tree in Figure 7 shows that occupants of open offices are more likely to belong to the dissatisfaction profiles concerned with acoustics, privacy, and space. A recent analysis of the CBE Occupant Survey (Kent et al. 2021) demonstrates the overwhelming influence of privacy and space on occupant dissatisfaction. It is reasonable to assume that shrinking work areas and removing acoustic and visual barriers will lead to a rise in complaints of this nature (De Croon et al. 2005; Keeling et al. 2015). Indeed, the most common rules for visual privacy, ease of interaction, and amount of space show a clear relationship between these dissatisfaction sources (Figure 8). Yet the prominence of such problems, along with the loss of personalization (Laurence et al. 2013), indicates that many designers are not adequately considering these when specifying open-plan layouts.

It is these challenges that have come to characterize the problems of open layout (Kim & de Dear 2013) and should therefore be the focus when mitigating sources of dissatisfaction in open-plan offices. Efforts to address this include architectural (Haapakangas et al. 2017), technological, and policy interventions (Babapour Chafi & Rolfö 2019). Aside from design interventions, one possible solution to compensate for a loss of privacy and space is to give occupants access to window views. Another CBE analysis found that occupants near a window were less likely to be dissatisfied, and similar results have been reported in laboratory experiments (Ko et al. 2020). There are also economic arguments for increasing window access (Turan et al. 2021). This is compelling evidence against the push by developers for deeper floorplates to maximize the floor area ratio. There may be increased facade and energy costs associated with providing occupants greater access to windows (and views), but this will likely reduce occupant dissatisfaction and potentially boost rental returns.

4.1.2 Acoustics

Acoustics remains the most prominent issue in contemporary office designs (Graham et al. 2021). Nearly half of all surveyed occupants in the CBE Occupant Survey selected acoustics as a dissatisfaction source. The three most common sources—people talking, people on their phones, and speech privacy—demonstrate that speech distraction and intelligibility is the biggest challenge in workspaces. This is in line with the extant literature on workplace acoustics (Haapakangas et al. 2017; Jahncke et al. 2013; Yadav et al. 2017). Active interventions might include acoustic absorption paneling (Passero & Zannin 2012), sound-masking (Hongisto et al. 2016), or introducing sound barriers such as furniture designed to provide a sense of acoustic and visual privacy. Policy solutions may aim to reduce noise sources by migrating some activities from voice to text (e.g. instant messaging) or moving others to dedicated meeting spaces or quiet rooms (Haapakangas et al. 2018). Others have suggested that solutions should reduce negative impacts while also striving to enhance auditory delights (Altomonte et al. 2020). The findings also show that equipment noise is very rarely the source of dissatisfaction. Many guidelines—e.g. ISO 3382-3 (ISO 2012) and ASHRAE’s Performance Measurement Protocols (ASHRAE 2012) suggest evaluating office acoustics using sound pressure measurements in an unoccupied space on the assumption that equipment noise is the issue. This suggests a need to modify compliance requirements for workspace acoustic assessments.

4.1.3 Thermal comfort

Thermal comfort in office buildings is a research topic that encompasses investigations of workplace satisfaction, thermal physiology, and HVAC system design. Yet satisfaction with the thermal environments in buildings rarely meets the arbitrary 80% acceptability targets found in standards (Arens et al. 2010; Li et al. 2019). Some of the most frequent dissatisfaction sources for thermal comfort from the survey—no access to the thermostat, different temperatures in the office, and a slow-responding HVAC system—all relate to a lack of control over the thermal environment. Two other common sources concern air movement, with an almost even number saying it is too high or too low. Another analysis of the same database found that gender and overcooling may explain some of the differences in air movement preference (Parkinson et al. 2021). In general, most of the problems in thermal comfort arise from the unfounded assumption that office environments should be tightly controlled around ‘optimum’ temperatures. Metanalyses of thermal comfort (Cheung et al. 2019) and worker performance (Porras-Salazar et al. 2021) provide overwhelming empirical evidence to disprove this practice.

The majority of the thermal dissatisfaction votes in the CBE Occupant Survey are attributable to five sources. Grouped into three challenges—different temperature preferences, perceived control, and air movement—these can all be directly addressed by personal comfort systems (Zhang et al. 2015) and personal comfort models (Kim et al. 2018). Such devices are often low-energy solutions such as desk fans and heated and cooled chairs that give occupants the power to modify their thermal environment to suit their preferences. Field studies have shown personal comfort systems increase thermal satisfaction compared with centralized HVAC and can sustain these high satisfaction levels across a wide range of temperatures (Zhang et al. 2015; Kim et al. 2019). Personal comfort models maximize the adaptability of the centralized or personalized systems to the preference of occupants.

4.1.4 Visual privacy and space

The two categories of visual privacy and amount of space concern issues that emerge with densification of office layouts. This design inevitably leads to shrinking personal work areas and an increased need for both acoustic and visual privacy. The findings of the association rule-mining (Figure 8) confirmed the connections between visual privacy (passers-by), amount of space (insufficient personal work area), and acoustics (people talking). This demonstrates that many open office designs have not found a good balance between increasing openness while providing sufficient privacy for occupants. This is being more closely examined as people return to offices following the Covid-19 pandemic. In the context of airborne transmission, greater distance between occupants (combined with increased ventilation, filtration, disinfection, and reduced densities) is one response to incentivizing workers to use the office again (Morawska et al. 2020). Possible interventions for better privacy and space include better ratios between open desk areas, private/focused work areas, improved office layout and meeting rooms. Architectural solutions may include additional storage space such as lockers or reorganizing the layout to reduce passers-by, such as that of cul-de-sac city planning. Policy solutions overlap with acoustics by discouraging conversations in open spaces.

4.1.5 Air quality

Smells from food is the most common complaint about the indoor air quality of offices. Many kitchens in contemporary office designs are positioned in the core of the floorplate. Kitchen extraction systems may be insufficiently sized to prevent food odors wafting into personal work areas in open offices. A complementary policy solution would be to discourage occupants from eating at their desk or in other common spaces. In addition, off-gassing from carpets and furniture appears to be a common problem despite this being a well-known challenge for indoor air quality (Weschler 2009). Careful product selection with low-volatile organic compound (VOC) materials may help address this issue (Wang et al. 2020).

Indoor air quality can be mitigated by increased ventilation rates (Sundell et al. 2011), improved filtration (Nazaroff 2004), or improved air distribution such as displacement ventilation and underfloor air distribution (Yang et al. 2019). There has been a flurry of activity on this topic following the Covid-19 pandemic (Morawska et al. 2021). Caution should be given, however, to the increased energy costs of higher fresh air rates.

Lastly, the ‘other’ category was the second most frequent complaint. This suggests the list of items shown to respondents does not cover commonly encountered air quality issues; future research efforts to better understand dissatisfaction sources for indoor air quality in office buildings are encouraged.

4.1.6 Lighting

A desire for more natural light was the most common complaint about lighting. This ties in with extant literature on occupant preferences (Galasiu & Veitch 2006) and emerging evidence showing the importance of window views in offices (Lottrup et al. 2015; Kent & Schiavon 2020; Ko et al. 2021). For artificial lighting, though, there is an almost equal number of occupants who experience too much light as those who think there is not enough light. Workspace environments should provide personalized lighting conditions that occupants can control and adapt to their needs (van Duijnhoven et al. 2021). Association rule-mining shown in Figure 8 found that those wanting more natural light also wanted more artificial light and found the light color undesirable. New lighting technologies with variable color temperatures and intensity to change with diurnal patterns to complement circadian rhythms are increasingly popular. Task lighting from desk lamps may also provide adequate light levels while offering personal control. On the other side, occupants experiencing too much light are the same who report reflections and shadows. One might assume that these complaints result from a lack of control over the lit environment, but that was not a common reason given by occupants. In both instances, there is an opportunity for novel facade designs to maximize the available natural light while also providing solutions for minimizing it where needed.

4.1.7 Ease of interaction

Occupants were concerned that noise from collegial interactions would impinge upon others in the space. This is related to the second most common complaint of limited casual spaces. These results suggest some offices do not have spaces to adequately facilitate work or social interactions. This also connects to visual privacy and amount of space, as well as the most common acoustic issue of people talking. There is a need for better configuration of space given the allocated areas for focused work, collaboration, and social areas may not be ideal for occupants in many offices. Additional policy interventions might help improve current practices in flexible offices (Babapour Chafi & Rolfö 2019).

4.2 Limitations

The key limitation of this work is the self-selection bias of the CBE Occupant Survey. Many of the organizations using POEs occupy premium-grade offices housing knowledge workers which is not representative of all occupants or buildings. Indeed, the overwhelming number of office buildings in the database from the US implies limited geographical and cultural contexts. Another challenge with optional surveys is non-response bias, which may skew responses towards occupants with a strong opinion about their workspace. While this may lead to underrepresentation of some response types, it is less problematic here as the focus was on dissatisfaction sources. Lastly, the CBE Occupant Survey was designed principally as an IEQ diagnostic tool. As such, it focuses on certain factors of the physical environment that form only part of any evaluation of workspace satisfaction. Future versions of the survey aim to expand the set of questions to include additional information about the occupants themselves (e.g. satisfaction with life, satisfaction with job, personality type, etc.), other factors relevant to work environments (e.g. agile designs, hybrid work, window views, etc.), and further evaluation of successful aspects of workspace designs.

5. Conclusions

A systematic analysis of dissatisfaction sources using post-occupancy evaluations (POEs) from over 600 office buildings was performed. The dataset came from branching questions in the Center for the Built Environment (CBE) Occupant Survey prompting occupants expressing dissatisfaction with an aspect of their workplace to nominate the reason for their dissatisfaction. Several themes emerged from the analysis that may be relevant in many contemporary workspace designs. First, acoustics and temperature are the main sources of dissatisfaction in office environments. These and other common challenges generally relate to the simultaneous reduction in control and personalization with increasingly open and densely populated layouts. Second, contextual factors influenced occupant dissatisfaction in ways that may not directly relate to the indoor environment. Office layout, time in building, and distance to a window significantly changed the likelihood of dissatisfaction. Third, these contextual factors appear to shape dissatisfaction profiles across different satisfaction items. Almost half of the respondents expressing dissatisfaction with some aspect of their workspace had more than one complaint, with acoustics, space, and privacy-related items co-occurring for many occupants. Lastly, there is a clear relationship between satisfaction with the environment and feeling supported by the workspace to complete work tasks.

This analysis of common sources of dissatisfaction could help to improve the design, operation, and maintenance of new workspaces and retrofit projects. Future research efforts should develop a greater understanding of the people occupying the space and the way they connect with the indoor environment.

Acknowledgements

The authors express their gratitude to Lindsay Graham for her support and leadership of the Center for the Built Environment (CBE) Occupant Survey; and to the Berkeley Research Impact Initiative (BRII), sponsored by the University of California—Berkeley Library, whose support made publication of this study possible.

Author contributions

G.B.: conceptualization, writing—review and editing. J.K.: conceptualization, writing—original draft, writing—review and editing. T.P.: conceptualization, data curation, formal analysis, funding acquisition, methodology, visualization, writing—original draft, writing—review and editing. S.S.: conceptualization, formal analysis, funding acquisition, methodology, writing—review and editing.

Competing interests

The Center for the Built Environment (CBE) at the University of California—Berkeley, with which the authors are affiliated, is advised, and funded in part, by many partners that represent a diversity of organizations from the building industry, including manufacturers, building owners, facility managers, contractors, architects, engineers, government agencies, and utilities.

Data availability

These data are not available for open access in compliance with the institutional review board (IRB) protocol approved for this study. Data summaries related to this paper may be requested from the authors.

Ethical approval

This study was approved by the institutional review board (IRB) at the University of California—Berkeley (number IRB2010-05-1550).

Funding

This research was (in part) funded by the industry consortium members of the Center for the Built Environment (CBE), University of California—Berkeley. T.P. and S.S. are partially supported by the Republic of Singapore’s National Research Foundation through a grant to the Berkeley Education Alliance for Research in Singapore (BEARS) for the Singapore–Berkeley Building Efficiency and Sustainability in the Tropics (SinBerBEST) Program. BEARS was established by the University of California—Berkeley as a center for intellectual excellence in research and education in Singapore. Publication was made possible in part by support from the Berkeley Research Impact Initiative (BRII) sponsored by the UC Berkeley Library.

References

  1. Altomonte, S., Allen, J., Bluyssen, P., Brager, G., Heschong, L., Loder, A., Schiavon, S., Veitch, J., Wang, L., & Wargocki, P. (2020). Ten questions concerning well-being in the built environment. Building and Environment, 106949. DOI: https://doi.org/10.1016/j.buildenv.2020.106949 

  2. Altomonte, S., Saadouni, S., Kent, M. G., & Schiavon, S. (2017). Satisfaction with indoor environmental quality in BREEAM and non-BREEAM certified office buildings. Architectural Science Review, 60(4), 343–355. DOI: https://doi.org/10.1080/00038628.2017.1336983 

  3. Altomonte, S., & Schiavon, S. (2013). Occupant satisfaction in LEED and non-LEED certified buildings. Building and Environment, 68, 66–76. DOI: https://doi.org/10.1016/j.buildenv.2013.06.008 

  4. Arens, E. A., Humphreys, M. A., de Dear, R. J., & Zhang, H. (2010). Are ‘class A’ temperature requirements realistic or desirable? Building and Environment, 45(1), 4–10. DOI: https://doi.org/10.1016/j.buildenv.2009.03.014 

  5. ASHRAE. (2012). Performance measurement protocols for commercial buildings: best practices guide. American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE). https://www.techstreet.com/standards/performance-measurement-protocols-for-commercial-buildings?product_id=1703581 

  6. Babapour Chafi, M., & Rolfö, L. (2019). Policies in activity-based flexible offices—I am sloppy with clean-desking. We don’t really know the rules. Ergonomics, 62(1), 1–20. DOI: https://doi.org/10.1080/00140139.2018.1516805 

  7. Bernstein, E. S., & Turban, S. (2018). The impact of the ‘open’ workspace on human collaboration. Philosophical Transactions of the Royal Society B: Biological Sciences, 373(1753), 20170239. DOI: https://doi.org/10.1098/rstb.2017.0239 

  8. Brager, G., & Baker, L. (2009). Occupant satisfaction in mixed-mode buildings. Building Research & Information, 37(4), 369–380. DOI: https://doi.org/10.1080/09613210902899785 

  9. Brennan, A., Chugh, J. S., & Kline, T. (2002). Traditional versus open office design: A longitudinal field study. Environment and Behavior, 34(3), 279–299. DOI: https://doi.org/10.1177/0013916502034003001 

  10. Burak Gunay, H., Shen, W., & Newsham, G. (2019). Data analytics to improve building performance: A critical review. Automation in Construction, 97, 96–109. DOI: https://doi.org/10.1016/j.autcon.2018.10.020 

  11. Cheung, T., Graham, L. T., & Schiavon, S. (2022). Impacts of life satisfaction, job satisfaction and the big five personality traits on satisfaction with the indoor environment. Building and Environment, 108783. DOI: https://doi.org/10.1016/j.buildenv.2022.108783 

  12. Cheung, T., Schiavon, S., Parkinson, T., Li, P., & Brager, G. (2019). Analysis of the accuracy on PMV-PPD model using the ASHRAE Global Thermal Comfort Database II. Building and Environment, 153, 205–217. DOI: https://doi.org/10.1016/j.buildenv.2019.01.055 

  13. Choi, J. H., Aziz, A., & Loftness, V. (2010). Investigation on the impacts of different genders and ages on satisfaction with thermal environments in office buildings. Building and Environment, 45(6), 1529–1535. DOI: https://doi.org/10.1016/j.buildenv.2010.01.004 

  14. De Been, I., & Beijer, M. (2014). The influence of office type on satisfaction and perceived productivity support. Journal of Facilities Management, 12(2), 142–157. DOI: https://doi.org/10.1108/JFM-02-2013-0011 

  15. De Croon, E., Sluiter, J., Kuijer, P. P., & Frings-Dresen, M. (2005). The effect of office concepts on worker health and performance: A systematic review of the literature. Ergonomics, 48(2), 119–134. DOI: https://doi.org/10.1080/00140130512331319409 

  16. Dutta, S., Gunay, H. B., & Bucking, S. (2020). Benchmarking operational performance of buildings by text mining tenant surveys. Science and Technology for the Built Environment, 1–20. DOI: https://doi.org/10.1080/23744731.2020.1851545 

  17. Dykes, C., & Baird, G. (2013). A review of questionnaire-based methods used for assessing and benchmarking indoor environmental quality. Intelligent Buildings International, 5(3), 135–149. DOI: https://doi.org/10.1080/17508975.2013.783457 

  18. Engelen, L., Chau, J., Young, S., Mackey, M., Jeyapalan, D., & Bauman, A. (2019). Is activity-based working impacting health, work performance and perceptions? A systematic review. Building Research & Information, 47(4), 468–479. DOI: https://doi.org/10.1080/09613218.2018.1440958 

  19. Fichman, M., & Levinthal, D. A. (1991). Honeymoons and the liability of adolescence: A new perspective on duration dependence in social and organizational relationships. Academy of Management Review, 16(2), 442–468. DOI: https://doi.org/10.5465/amr.1991.4278962 

  20. Frontczak, M., Schiavon, S., Goins, J., Arens, E. A., Zhang, H., & Wargocki, P. (2012). Quantitative relationships between occupant satisfaction and satisfaction aspects of indoor environmental quality and building design. Indoor Air, 22(2), 119–131. DOI: https://doi.org/10.1111/j.1600-0668.2011.00745.x 

  21. Galasiu, A. D., & Veitch, J. A. (2006). Occupant preferences and satisfaction with the luminous environment and control systems in daylit offices: A literature review. Energy and Buildings, 38(7), 728–742. DOI: https://doi.org/10.1016/j.enbuild.2006.03.001 

  22. Goins, J., & Moezzi, M. (2013). Linking occupant complaints to building performance. Building Research & Information, 41(3), 361–372. DOI: https://doi.org/10.1080/09613218.2013.763714 

  23. Graham, L. T., Parkinson, T., & Schiavon, S. (2021). Lessons learned from 20 years of CBE’s occupant surveys. Buildings & Cities, 2(1), 166–184. DOI: https://doi.org/10.5334/bc.76 

  24. Haapakangas, A., Hongisto, V., Eerola, M., & Kuusisto, T. (2017). Distraction distance and perceived disturbance by noise—An analysis of 21 open-plan offices. Journal of the Acoustical Society of America, 141(1), 127–136. DOI: https://doi.org/10.1121/1.4973690 

  25. Haapakangas, A., Hongisto, V., Varjo, J., & Lahtinen, M. (2018). Benefits of quiet workspaces in open-plan offices—Evidence from two office relocations. Journal of Environmental Psychology, 56, 63–75. DOI: https://doi.org/10.1016/j.jenvp.2018.03.003 

  26. Hahsler, M., Buchta, C., Gruen, B., & Hornik, K. (2021). arules: Mining Association rules and frequent itemsets. R package version 1.6-6. https://CRAN.R-project.org/package=arules 

  27. Heinzerling, D., Schiavon, S., Webster, T., & Arens, E. (2013). Indoor environmental quality assessment models: A literature review and a proposed weighting and classification scheme. Building and Environment, 70, 210–222. DOI: https://doi.org/10.1016/j.buildenv.2013.08.027 

  28. Hongisto, V., Varjo, J., Leppämäki, H., Oliva, D., & Hyönä, J. (2016). Work performance in private office rooms: The effects of sound insulation and sound masking. Building and Environment, 104, 263–274. DOI: https://doi.org/10.1016/j.buildenv.2016.04.022 

  29. ISO. (2012). ISO Standard 3382-3: Acoustics—Measurement of Room Acoustic Parameters—Part 3: Open Plan Offices. International Organization for Standardization (ISO). https://www.iso.org/standard/46520.html 

  30. Jahncke, H., Hongisto, V., & Virjonen, P. (2013). Cognitive performance during irrelevant speech: Effects of speech intelligibility and office-task characteristics. Applied Acoustics, 74(3), 307–316. DOI: https://doi.org/10.1016/j.apacoust.2012.08.007 

  31. Karmann, C., Schiavon, S., Graham, L. T., Raftery, P., & Bauman, F. (2017). Comparing temperature and acoustic satisfaction in 60 radiant and all-air buildings. Building and Environment, 126, 431–441. DOI: https://doi.org/10.1016/j.buildenv.2017.10.024 

  32. Keeling, T., Clements-Croome, D., & Roesch, E. (2015). The effect of agile workspace and remote working on experiences of privacy, crowding and satisfaction. Buildings, 5(3), 880–898. DOI: https://doi.org/10.3390/buildings5030880 

  33. Kent, M., Parkinson, T., Kim, J., & Schiavon, S. (2021). A data-driven analysis of occupant workspace dissatisfaction. Building and Environment, 205, 108270. DOI: https://doi.org/10.1016/j.buildenv.2021.108270 

  34. Kent, M., & Schiavon, S. (2020). Evaluation of the effect of landscape distance seen in window views on visual satisfaction. Building and Environment, 183, 107160. DOI: https://doi.org/10.1016/j.buildenv.2020.107160 

  35. Kim, J., Bauman, F., Raftery, P., Arens, E., Zhang, H., Fierro, G., Andersen, M., & Culler, D. (2019). Occupant comfort and behavior: High-resolution data from a 6-month field study of personal comfort systems with 37 real office workers. Building and Environment, 148, 348–360. DOI: https://doi.org/10.1016/j.buildenv.2018.11.012 

  36. Kim, J., Candido, C., Thomas, L., & de Dear, R. (2016). Desk ownership in the workplace: The effect of non-territorial working on employee workplace satisfaction, perceived productivity and health. Building and Environment, 103, 203–214. DOI: https://doi.org/10.1016/j.buildenv.2016.04.015 

  37. Kim, J., & de Dear, R. (2012). Impact of different building ventilation modes on occupant expectations of the main IEQ factors. Building and Environment, 57, 184–193. DOI: https://doi.org/10.1016/j.buildenv.2012.05.003 

  38. Kim, J., & de Dear, R. (2013). Workspace satisfaction: The privacy–communication trade-off in open-plan offices. Journal of Environmental Psychology, 36(0), 18–26. DOI: https://doi.org/10.1016/j.jenvp.2013.06.007 

  39. Kim, J., de Dear, R., Cândido, C., Zhang, H., & Arens, E. (2013). Gender differences in office occupant perception of indoor environmental quality (IEQ). Building and Environment, 70, 245–256. DOI: https://doi.org/10.1016/j.buildenv.2013.08.022 

  40. Kim, J., Schiavon, S., & Brager, G. (2018). Personal comfort models—A new paradigm in thermal comfort for occupant-centric environmental control. Building and Environment, 132, 114–124. DOI: https://doi.org/10.1016/j.buildenv.2018.01.023 

  41. Ko, W. H., Kent, M. G., Schiavon, S., Levitt, B., & Betti, G. (2021). A window view quality assessment framework. LEUKOS. DOI: https://doi.org/10.1080/15502724.2021.1965889 

  42. Ko, W. H., Schiavon, S., Zhang, H., Graham, L. T., Brager, G., Mauss, I., & Lin, Y.-W. (2020). The impact of a view from a window on thermal comfort, emotion, and cognitive performance. Building and Environment, 175, 106779. DOI: https://doi.org/10.1016/j.buildenv.2020.106779 

  43. Laurence, G. A., Fried, Y., & Slowik, L. H. (2013). ‘My space’: A moderated mediation model of the effect of architectural and experienced privacy and workspace personalization on emotional exhaustion at work. Journal of Environmental Psychology, 36, 144–152. DOI: https://doi.org/10.1016/j.jenvp.2013.07.011 

  44. Leder, S., Newsham, G. R., Veitch, J. A., Mancini, S., & Charles, K. E. (2016). Effects of office environment on employee satisfaction: a new analysis. Building Research & Information, 44(1), 34–50. DOI: https://doi.org/10.1080/09613218.2014.1003176 

  45. Li, P., Froese, T. M., & Brager, G. (2018). Post-occupancy evaluation: State-of-the-art analysis and state-of-the-practice review. Building and Environment, 133, 187–202. DOI: https://doi.org/10.1016/j.buildenv.2018.02.024 

  46. Li, P., Parkinson, T., Brager, G., Schiavon, S., Cheung, T. C. T., & Froese, T. (2019). A data-driven approach to defining acceptable temperature ranges in buildings. Building and Environment, 153, 302–312. DOI: https://doi.org/10.1016/j.buildenv.2019.02.020 

  47. Lottrup, L., Stigsdotter, U. K., Meilby, H., & Claudi, A. G. (2015). The workplace window view: A determinant of office workers’ work ability and job satisfaction. Landscape Research, 40(1), 57–75. DOI: https://doi.org/10.1080/01426397.2013.829806 

  48. Maechler, M., Rousseeuw, P., Struyf, A., Hubert, M., Hornik, K. (2021). cluster: Cluster Analysis Basics and Extensions. R package version 2.1.2. https://CRAN.R-project.org/package=cluster 

  49. McArthur, J. J., Shahbazi, N., Fok, R., Raghubar, C., Bortoluzzi, B., & An, A. (2018). Machine learning and BIM visualization for maintenance issue classification and enhanced data collection. Advanced Engineering Informatics, 38, 101–112. DOI: https://doi.org/10.1016/j.aei.2018.06.007 

  50. Moezzi, M., & Goins, J. (2011). Text mining for occupant perspectives on the physical workplace. Building Research & Information, 39(2), 169–182. DOI: https://doi.org/10.1080/09613218.2011.556008 

  51. Morawska, L., Allen, J., Bahnfleth, W., Bluyssen, P. M., Boerstra, A., Buonanno, G., Cao, J., Dancer, S. J., Floto, A., Franchimon, F., Greenhalgh, T., Haworth, C., Hogeling, J., Isaxon, C., Jimenez, J. L., Kurnitski, J., Li, Y., Loomans, M., Marks, G., … Yao, M. (2021). A paradigm shift to combat indoor respiratory infection. Science, 372(6543), 689–691. DOI: https://doi.org/10.1126/science.abg2025 

  52. Morawska, L., Tang, J. W., Bahnfleth, W., Bluyssen, P. M., Boerstra, A., Buonanno, G., Cao, J., Dancer, S., Floto, A., Franchimon, F., Haworth, C., Hogeling, J., Isaxon, C., Jimenez, J. L., Kurnitski, J., Li, Y., Loomans, M., Marks, G., Marr, L. C., … Yao, M. (2020). How can airborne transmission of COVID-19 indoors be minimised? Environment International, 142, 105832. DOI: https://doi.org/10.1016/j.envint.2020.105832 

  53. Nazaroff, W. W. (2004). Indoor particle dynamics. Indoor Air, 14(s7), 175–83. DOI: https://doi.org/10.1111/j.1600-0668.2004.00286.x 

  54. Oldham, G. R., & Brass, D. J. (1979). Employee reactions to an open-plan office: A naturally occurring quasi-experiment. Administrative Science Quarterly, 24(2), 267. DOI: https://doi.org/10.2307/2392497 

  55. Parkinson, T., Schiavon, S., de Dear, R., & Brager, G. (2021). Overcooling of offices reveals gender inequity in thermal comfort. Scientific Reports, 11(1), 23684. DOI: https://doi.org/10.1038/s41598-021-03121-1 

  56. Passero, C. R. M., & Zannin, P. H. T. (2012). Acoustic evaluation and adjustment of an open-plan office through architectural design and noise control. Applied Ergonomics, 43(6), 1066–1071. DOI: https://doi.org/10.1016/j.apergo.2012.03.007 

  57. Porras-Salazar, J. A., Schiavon, S., Wargocki, P., Cheung, T., & Tham, K. W. (2021). Meta-analysis of 35 studies examining the effect of indoor temperature on office work performance. Building and Environment, 203, 108037. DOI: https://doi.org/10.1016/j.buildenv.2021.108037 

  58. Preiser, W. F. E., Rabinowitz, H. Z., & White, E. T. (2015). Post-occupancy evaluation. Routledge. DOI: https://doi.org/10.4324/9781315713519 

  59. Rousseeuw, P. J. (1987). Silhouettes: A graphical aid to the interpretation and validation of cluster analysis. Journal of Computational and Applied Mathematics, 20, 53–65. DOI: https://doi.org/10.1016/0377-0427(87)90125-7 

  60. Sundell, J., Levin, H., Nazaroff, W. W., Cain, W. S., Fisk, W. J., Grimsrud, D. T., Gyntelberg, F., Li, Y., Persily, A. K., Pickering, A. C., Samet, J. M., Spengler, J. D., Taylor, S. T., & Weschler, C. J. (2011). Ventilation rates and health: Multidisciplinary review of the scientific literature. Indoor Air, 21(3), 191–204. DOI: https://doi.org/10.1111/j.1600-0668.2010.00703.x 

  61. Therneau, T., & Atkinson, B. (2019). rpart: Recursive Partitioning and Regression Trees. R package version 4.1-15. https://CRAN.R-project.org/package=rpart 

  62. Turan, I., Chegut, A., Fink, D., & Reinhart, C. (2021). Development of view potential metrics and the financial impact of views on office rents. Landscape and Urban Planning, 215, 104193. DOI: https://doi.org/10.1016/j.landurbplan.2021.104193 

  63. van Duijnhoven, J., Aarts, M. P. J., van den Heuvel, E. R., & Kort, H. S. M. (2021). The identification of variables influencing personal lighting conditions of office workers. Lighting Research & Technology, 53(6), 527–541. DOI: https://doi.org/10.1177/1477153520976950 

  64. Vischer, J. C. (2008). Towards an environmental psychology of workspace: How people are affected by environments for work. Architectural Science Review, 51(2), 97–108. DOI: https://doi.org/10.3763/asre.2008.5114 

  65. Wang, C., Collins, D. B., Arata, C., Goldstein, A. H., Mattila, J. M., Farmer, D. K., Ampollini, L., et al. (2020). Surface reservoirs dominate dynamic gas-surface partitioning of many indoor air constituents. Science Advances, 6(8). DOI: https://doi.org/10.1126/sciadv.aay8973 

  66. Weschler, C. J. (2009). Changes in indoor pollutants since the 1950s. Atmospheric Environment, 43(1), 153–169. DOI: https://doi.org/10.1016/j.atmosenv.2008.09.044 

  67. Wickham, H., Averick, M., Bryan, J., Chang, W., D’Agostino McGowan, L., …, Yutani, H. (2019). Welcome to the tidyverse. Journal of Open Source Software, 4(43), 1686. DOI: https://doi.org/10.21105/joss.01686 

  68. Yadav, M., Kim, J., Cabrera, D., & de Dear, R. (2017). Auditory distraction in open-plan office environments: The effect of multi-talker acoustics. Applied Acoustics, 126, 68–80. DOI: https://doi.org/10.1016/j.apacoust.2017.05.011 

  69. Yang, B., Melikov, A. K., Kabanshi, A., Zhang, C., Bauman, F. S., … Lin, Z. (2019). A review of advanced air distribution methods—Theory, practice, limitations and solutions. Energy and Buildings, 202, 109359. DOI: https://doi.org/10.1016/j.enbuild.2019.109359 

  70. Zagreus, L., Huizenga, C., Arens, E., & Lehrer, D. (2004). Listening to the occupants: A web-based indoor environmental quality survey. Indoor Air, 14(8: Suppl.), 65–74. DOI: https://doi.org/10.1111/j.1600-0668.2004.00301.x 

  71. Zhang, H., Arens, E. A., & Zhai, Y. (2015). A review of the corrective power of personal comfort systems in non-neutral ambient environments. Building and Environment, 91, 15–41. DOI: https://doi.org/10.1016/j.buildenv.2015.03.013 

  72. Zimmerman, A., & Martin, M. (2001). Post-occupancy evaluation: benefits and barriers. Building Research & Information, 29(2), 168–174. DOI: https://doi.org/10.1080/09613210010016857 

A1. Appendix A

Example of the branching question format in the Center for the Built Environment (CBE) Occupant Survey
Figure A1 

Example of the branching question format in the Center for the Built Environment (CBE) Occupant Survey.

Note: Respondents who are dissatisfied (any vote below ‘neither’) are shown a checklist of potential reasons for their dissatisfaction, i.e. dissatisfaction sources.

Frequency of the 20 most common dissatisfaction sources across the entire dataset
Figure A2 

Frequency of the 20 most common dissatisfaction sources across the entire dataset.

Note: Colors indicate the category of the dissatisfaction source.

comments powered by Disqus